SRA

Comparison of Solexa and Affy expression in two human tissues. Ultra high-throughput sequencing has been used recently as an alternative to microarrays for genotyping, analysis of methylation patterns and identification of transcription factor binding sites. Here, we describe an application of the Solexa sequencing platform to the detection and comparison of mRNA expression levels. Our goals were to estimate technical variance associated with Solexa sequencing and to assess its ability to identify differentially expressed genes. To do so, we estimated gene expression differences between liver and kidney RNA samples using multiple sequencing replicates, and compared the sequencing data to results obtained from arrays, using the same RNA samples. We show that the sequencing data are highly replicable, with remarkably little technical variation. Indeed, our analyses suggest it may be sufficient to sequence each mRNA sample only once (i.e., using one lane) to accurately estimate gene expression levels. Moreover, our data suggests that Solexa sequencing is comparable, if not superior, to arrays in facilitating the identiffication of differ­entially expressed genes, while allowing for additional analyses such as detection of low-expressed genes, alternative splice variants, and novel transcripts. Based on our observations, we propose an empirical protocol as well as a statistical framework for the analysis of gene expression using ultra high-throughput sequencing technology. We extracted total RNA from liver and kidney samples of a single human male, purified the poly-A mRNA and sheared it prior to cDNA synthesis. The cDNA was then processed into a library of template molecules suitable for sequencing on the Illumina Genome Analyzer. To assess technical variance within and between runs, we sequenced each sample seven times, split across two runs of the machine. To investigate the effects of cDNA concentration, two different cDNA concentrations were used: 3 pM (five lanes per sample) and 1.5 pM (two lanes per sample). Raw 36bp reads were trimmed to 32bp to select high-quality data from the Solexa instrument